schedule_sampling: Define sampling events for a given set of populations

In slendr: A Simulation Framework for Spatiotemporal Population Genetics

Define sampling events for a given set of populations

Description

Schedule sampling events at specified times and, optionally, a given set of locations on a landscape

Usage

schedule_sampling(model, times, ..., locations = NULL, strict = FALSE)

Arguments

model	Object of the class slendr_model
times	Integer vector of times (in model time units) at which to schedule remembering of individuals in the tree-sequence
...	Lists of two elements (slendr_pop population object-<number of individuals to sample), representing from which populations should how many individuals be remembered at times given by times
locations	List of vector pairs, defining two-dimensional coordinates of locations at which the closest number of individuals from given populations should be sampled. If NULL (the default), individuals will be sampled randomly throughout their spatial boundary.
strict	Should any occurence of a population not being present at a given time result in an error? Default is FALSE, meaning that invalid sampling times for any populations will be quietly ignored.

Details

If both times and locations are given, the the sampling will be scheduled on each specified location in each given time-point. Note that for the time-being, in the interest of simplicity, no sanity checks are performed on the locations given except the restriction that the sampling points must fall within the bounding box around the simulated world map. Other than that, slendr will simply instruct its SLiM backend script to sample individuals as close to the sampling points given as possible, regardless of whether those points lie within a population spatial boundary at that particular moment of time.

Optionally, a name of a single sample from a population can be given, which will then replace the generic format of "\{population\}_\{number\}" name. See example for more detail.

Value

Data frame with three columns: time of sampling, population to sample from, how many individuals to sample

Examples

init_env()

# load an example model with an already simulated tree sequence
path <- system.file("extdata/models/introgression", package = "slendr")
model <- read_model(path)

# afr, eur, and nea objects would normally be created before slendr model
# compilation, but here we take them out of the model object already compiled for
# this example (in a standard slendr simulation pipeline, this wouldn't be necessary)
afr <- model$populations[["AFR"]]
eur <- model$populations[["EUR"]]
nea <- model$populations[["NEA"]]

# schedule the recording of 10 African and 100 European individuals from a
# given model at 20 ky, 10 ky, 5ky ago and at present-day (time 0)
schedule_amh <- schedule_sampling(
  model, times = c(20000, 10000, 5000, 0),
  list(afr, 10), list(eur, 100)
)
# schedule the recording of the Vindija Neanderthal genome
schedule_nea <- schedule_sampling(model, times = 40000, list(nea, 1, "Vindija"))

# the result of `schedule_sampling` is a simple data frame (note that the locations
# of sampling locations have `NA` values because the model is non-spatial)
schedule <- rbind(schedule_amh, schedule_nea)
schedule

# simulate a tree sequence
ts <- msprime(model, sequence_length = 1e6, recombination_rate = 1e-8, samples = schedule)

# inspect the recorded table of samples
ts_samples(ts)

slendr documentation built on Aug. 8, 2025, 7:14 p.m.